SlideShare a Scribd company logo
1 of 1
Download to read offline
Faster Hamiltonian Monte Carlo by Learning
Leapfrog Scale
Changye Wu, Julien Stoehr and Christian P. Robert.
Université Paris-Dauphine, Université PSL, CNRS, CEREMADE, 75016 PARIS, FRANCE
Abstract
Hamiltonian Monte Carlo samplers have become standard algo-
rithms for MCMC implementations, as opposed to more basic
versions, but they still require some amount of tuning and cali-
bration. Exploiting the U-turn criterion of the NUTS algorithm
[2], we propose a version of HMC that relies on the distribution
of the integration time of the associated leapfrog integrator.
Using in addition the primal-dual averaging method for tuning
the step size of the integrator, we achieve an essentially calibra-
tion free version of HMC. When compared with the original
NUTS on benchmarks, this algorithm exhibits a significantly
improved efficiency.
Hamiltonian Monte Carlo (HMC, [3])
Consider a density π on Θ ⊂ Rd
with respect to the Lebesgue
measure,
π(θ) ∝ exp{−U(θ)}, where U ∈ C1
(Θ).
Aim: generate a Markov chain (θ1, . . . , θN) with invariant dis-
tribution π to estimate, for some function h, functionals with
respect to π,
1
N
N
n=1
h(θn)
a.s.
−→
N→+∞ Θ
h(θ)π(dθ).
Principle: sample from an augmented target distribution
π(θ, v) = π(θ)N(v | 0, M) ∝ exp {−H(θ, v)} .
• auxiliary variable v ∈ Rd
referred to as momentum variable
as opposed to θ referred to as position,
• marginal chain in θ is the distribution of interest.
Hamiltonian dynamics: generating proposals for (θ, v) based
on



dθ
dt
=
∂H
∂v
= M−1
v
dv
dt
= −
∂H
∂θ
= − U(θ).
⊕ leaves π invariant, allows
large moves,
requires the solution flow
to the differential equa-
tions.
(θ, v)
(θ , v )
Leapfrog integrator: second order symplectic integrator
which yields an approximate solution flow by iterating the fol-
lowing procedure from (θ0, v0) = (θ, v)



r = vn − /2 U(θn),
θn+1 = θn + M−1
r,
vn+1 = r − /2 U(θn + 1).
• : a discretisation time-step.
• L: a number of leapfrog steps
solution at a time t = L
(θ, v)
(θ , v )
This scheme does no longer leave the measure π invariant!
Correction: an accept-reject step is introduced. A transition
from (θ, v) to proposal θ , −v is accepted with probability
ρ θ, v, θ , v = 1 ∧ exp H(θ, v) − H θ , −v .
Pros & cons:
⊕ The algorithm theoretically benefits from a fast explo-
ration of the parameter space by accepting large transi-
tions with high probability.
High sensitivity to hand-tuned parameters, namely the
step size of the discretisation scheme, the number of
steps L of the integrator, and the covariance matrix M.
The No-U-Turn Sampler (NUTS, [2])
Idea: version of HMC sampler that eliminates the need to
specify the number L by adaptively choosing the locally largest
value at each iteration of the algorithm.
How? Doubling the leapfrog path, either forward or back-
ward with equal probability, until the backward and the for-
ward end points of the path, (θ−
, v−
) and (θ+
, v+
), satisfy
(θ+
− θ−
) · M−1
v−
< 0
or
(θ+
− θ−
) · M−1
v+
< 0.
(θ, v)
(θ+
, v+
)
(θ−
, v−
)
Proposal: sampling along the generated trajectory
• slice sampling [2]
• multinomial sampling ([1], version implemented in
Stan).
What about ? Tuned via primal-dual averaging [4], by aim-
ing at a targeted acceptance probability δ0 ∈ (0, 1).
Numerical experiment: Susceptible-Infected-Recovered Model (SIR)
SIR: model used to represent disease transmission, for epidemics like cholera, within a population
ηk ∼ Poisson(ytk−1,1 − ytk,1) and ˆBk ∼ LN(ytk,4, 0.152
), where



k = 1, · · · , Nt = 20,
tk = 7k.
The dynamic of yt ∈ R4
is
dyt,1
dt
= −
βyt,4
yt,4 + κ0
yt,1,
dyt,2
dt
=
βyt,4
yt,4 + κ0
yt,1 − γyt,2,
dyt,3
dt
= γyt,3,
dyt,4
dt
= ξyt,2 − φyt,4,
Interpretation:
• yt,1, yt,2, yt,3 : no. of susceptible, infected, and recovered people
within the community.
• yt,4: concentration of the virus in the water reservoir.
• ηk: size of the pop. that becomes infected during [tk−1, tk].
Assumptions:
• the size of the population yt,1 +yt,2 +yt,3 is constant,
• β ∼ C(0, 2.5), γ ∼ C(0, 1),
ξ ∼ C(0, 25), φ ∼ C(0, 1),
Observed dataset: https://github.com/stan-dev/stat_comp_benchmarks/tree/master/benchmarks/sir.
q
q
qq
qq
qqq
qqqq
q
qqqqq
q
q
q
qq
q
q
q
q
qqqqqq
q
q
q
qq
q
qq
q
q
q
q
qqq
q
qqq
q
qqq
q
qq
q
qq
q
qqqqq
qq
q
q
q
q
q
q
qq
q
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
qqq
q
qqq
qqq
q
q
q
q
qqqqqqq
qqq q
qqqq
qq
q
qq
q
q
qqq
qq
q
qqqqqq
q
qqq
q
q
q
qq
q
qqq
q
qq qq
qqq
q
q
q
qqqqqqq
q
q
q
qqqq
q
qqq
q
qqq
q
qqq
qqqqqq qqqqqqqqqqqqqqqqqq
q
q
q
q
qqqq
q
qqqq
qq
q
qqq
q
q
q
q
q
q
qq
q
q
q
qq
q
q
qqq
q
qqqqq
q
q
qq
q
q
q
q
q
qqqqq
qq
q
q
q
q
q
qq
q
q
q
q
qq
q
q
q
q
q
qq
q
qq
qq
q
qq
qq
q
q
qqq
qq
qq
q
q
q
q
qqq
qqqq
q
q
q
q
qq
qqqq
q
qq
q
qq
q
q
q
q
qqq
qq
q
q
q
qqq
q
q
q
q
q
q
q
q
qqq
q
qqq
q
q
q
qq
q
q
qqq
q
q
qqq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
qq
q
q
q
qqq
qqq
qqq
q
qq
q
q
qq
q
q
q
q
qq
q
q
q
q
q
qq
q
q
qq
qq
q
q
q
q
q
q
q
q
q
q
q
qqq
q
qq qq
q
qq
q
q
q
q
qqqq
q
q
q
q
q
q
q
q
qq
qq
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
qq
q
qq
q
q
q
qqq
q
qq
q
q
q
q
q
qqq
q
q
qqq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
qq
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
qq
q
q
q
q
qq
q
q
q
q
q
q
q
qq
qq
q
q
qq
q
q
q
q
qqq
qq
q
q
q
q
q
q
NUTS eHMC
0.6 0.7 0.8 0.9 0.6 0.7 0.8 0.9
0.00
0.01
0.02
ESJDpergradient
qq
q
q
q
q
q
q
q
qq
q
q
qq
q
q
qq
q
qqq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
qq
q
q
qqq
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
qq
qq
q
q
qqq
q
qq
qqq
qqqq
q
q
q
q
q
q
qqqqq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qqqqq
q
q
q
qq
q
q
q
q
q
q
qq
qq
qqq
q
q
q
qq
q
qq
qq
q
qq
q
q
q
q
qq
q
q
qq
q
qq
q
qqq
q
q
qq
q
q
q
q
qq
q
qq
q
qqqqq
q
q
qq
q
q
qqq
q
q
qq
q
qq
q
q
qq
q
q
q
qq
q
q
q
qqq
q
q
q
q
q
q
qqq
q
qq
q
q
qq
q
q
q
qqq
q
q
qq
q
q
q
q
q
q
qqq
q
q
q
q
q
q
q
q
qq
qq
q
q
q
q
q
q
q
qqq
q
qqq
qqq
q
q
q
qq
q
qq
q
q
q
qq
q
q
q
q
q
q
q
qq
q
q
q
q
qq
q
q
q
q
q
q
qqqqq
q
q
q
q
qq
q
qq
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
qq
qq
q
qq
q
q
q
q
qqq
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
qq
q
q
q
q
qqqq
q
q
q
q
q
q
q
q
q
q
q
qqqq
q
q
q
q
q
q
qq
q
q
q
qq
q
q
q
q
q
q
q
q
qq
qq
qq
q
qqqqq
q
q
q
qq
q
q
q
q
q
q
q
qq
qq
q
q
q
qq
q
q
q
q
q
qq
q
qq
q
qq
q
qqq
q
q
qqq
q
qqq
qq
q
q
qq
q
q
qq
qq
qq
q
q
q
q
q
q
q
q
q
qq
q
qq
qq
q
qqq
q
q
q
q
q
q
qqq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qqq
q
qq
q
q
qqq
q
q
q
q
qqq
q
qq
q
q
q
q
q
q
q
q
qq
qq
q
q
q
q
qqqqq
q
q
q
qq
q
qqq
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
qq
q
q
qq
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
qqqq
q
qq
q
q
q
q
q
q
qqq
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
qqq
q
q
q
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qqq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q q
q
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
qqq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
qq
q
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
qq
qq
q
q
q
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
qq
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qqq
q
qq
q
q
q
qq
q
q
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
qqqq
q
q
q
q
q
qqq
q
qq
q
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
qq
qq
q
qqq
qq
qq
qqqqqqqq
q
q
q
q
q
qq
q
qq
q
q
q
q
q
q
q
q
q
q
q
q
qqqq
q
q
q
qqqq
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
qqq
q
q
q
qqq
qq
qq
q
q
q
q
q
q
qqq
q
q
q
q
qq
q
q
qqq
q
q
qq
qqq
q
q
qq
qq
q
qqqqq
q
q
q
q
q
qqq
q
q
q
q
q
q
qq
qq
q
q
q
q
q
qq
q
qqqqq
q
q
q
q
q
q
q
qq
q
qq
q
q
qq
q
q
q
qqq
q
q
qq
q
q
qq
q
q
qqq
q
q
q
q
q
q
q
q
qqqq
q
q
q
qq
q
q
q
qq
q
q
qqqqq
q
q
q
qq
q
q
qqq
q
qq
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
qq
q
q
qq
q
q
q
q
q
q
q
q
q
q
qq
q
q
qqq
q
q
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qqq
q
qq
qq
q
q
q
q
q
q
qq
q
q
q
q
q
qq
q
q
q
q
q qqq
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
qq
q
q
q
q
qq
qq
q
q
qq
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
qqqq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qqq
q
q
q
qq
q
qq
q
qq
q
q
q
q
qqqq
q
q
q
q
q
q
q
q
q
qqq
qq
qq
q
q
q
q
q
q
q
q qq
q
qq
qq
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
q
qqq
q
q
q
q
q
qq
qq
q
q
q
qq qq
q
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
qqq
qqq
q
q
q
qqqqq
qq
qq
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
qq
q
q
q
q
q
q
qq
q
q
q
q
qqq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
qq
q
q
qq
q
q
q
q
q
qq
q
q
q
q
q
q
q
qq
qq
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qqq
q
q
q
q
q
q
q
qq
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
q
qq
q
q
q
q
q
qqq
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
qq
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
qq
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
qqq
q
q
q
q
q
qq
q
q
q
q
q
q
qq
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
qq
q
q
q
qqq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
qq
q
q
q
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
qq
q
q
q
q
q
q
qq
q
q
q
qq
q
q
q
q
q
qq
q
q
q
qq
q
q
q
qq
q
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
q
q
q
q
q
q
q
qq
qq
q
q
q
q
q
q
q
qq
q
q
q
q
q
q
q
qq
qq
q
qq
q
q
q
q
q
q
q
q
q
NUTS (xi) eHMC (xi) NUTS (phi) eHMC (phi)
NUTS (beta) eHMC (beta) NUTS (gamma) eHMC (gamma)
0.6 0.7 0.8 0.9 0.6 0.7 0.8 0.9 0.6 0.7 0.8 0.9 0.6 0.7 0.8 0.9
0.005
0.010
0.015
0.020
0.025
0.005
0.010
0.015
0.020
0.025
Targeted acceptance probability δ
ESSpergradient
Empirical HMC (eHMC, [5])
Longest batch associated with (θ, v, ):
L (θ, v) = inf ∈ N (θ − θ) · M−1
v < 0 ,
where (θ , v ) is the value of the pair after iterations of the
leapfrog integrator.
Learning leapfrog scale: tuning phase with the optimised
step size and an initial number of leapfrog steps L0. At each
iteration, one
1. iterates L0 leapfrog steps to generate the next state of the
Markov chain.
2. computes the longest batch for the current state of the chain.
Output: empirical distribution of the longest batches
ˆPL =
1
K
K−1
k=0
δ L (θk, v(k+1)
) .
eHMC: randomly pick a number of leapfrog steps accord-
ing to the empirical distribution ˆPL at each iteration of HMC
algorithm.
⇒ valid: the resulting transition kernel can be seen as a com-
position of multiple Markov kernels attached to the same
stationary distribution π.
References
[1] M. Betancourt A conceptual introduction to Hamiltonian Monte Carlo. arXiv preprint arXiv:1701.02434, 2017.
[2] M. D. Hoffmand and A. Gelman. The No-U-Turn Sampler: adaptively setting path lengths in Hamiltonian Monte Carlo. Journal of Machine Learning Research, 15(1):1593–1623, 2014.
[3] R. M. Neal. MCMC using Hamiltonian dynamics. Handbook of Markov Chain Monte Carlo, Chapter 5, Chapman & Hall / CRC Press, 2011.
[4] Y. Nesterov. Primal-dual subgradient methods for convex problems. Mathematical Programming, 120(1):221–259, 2009.
[5] C. Wu, J. Stoehr and C. P. Robert. Faster Hamiltonian Monte Carlo by Learning Leapfrog Scale. arXiv preprint arXiv:1810.04449v1, 2018.

More Related Content

What's hot

Mark Girolami's Read Paper 2010
Mark Girolami's Read Paper 2010Mark Girolami's Read Paper 2010
Mark Girolami's Read Paper 2010Christian Robert
 
Rao-Blackwellisation schemes for accelerating Metropolis-Hastings algorithms
Rao-Blackwellisation schemes for accelerating Metropolis-Hastings algorithmsRao-Blackwellisation schemes for accelerating Metropolis-Hastings algorithms
Rao-Blackwellisation schemes for accelerating Metropolis-Hastings algorithmsChristian Robert
 
Approximate Bayesian Computation with Quasi-Likelihoods
Approximate Bayesian Computation with Quasi-LikelihoodsApproximate Bayesian Computation with Quasi-Likelihoods
Approximate Bayesian Computation with Quasi-LikelihoodsStefano Cabras
 
Bayesian hybrid variable selection under generalized linear models
Bayesian hybrid variable selection under generalized linear modelsBayesian hybrid variable selection under generalized linear models
Bayesian hybrid variable selection under generalized linear modelsCaleb (Shiqiang) Jin
 
Inference in generative models using the Wasserstein distance [[INI]
Inference in generative models using the Wasserstein distance [[INI]Inference in generative models using the Wasserstein distance [[INI]
Inference in generative models using the Wasserstein distance [[INI]Christian Robert
 
Maximum likelihood estimation of regularisation parameters in inverse problem...
Maximum likelihood estimation of regularisation parameters in inverse problem...Maximum likelihood estimation of regularisation parameters in inverse problem...
Maximum likelihood estimation of regularisation parameters in inverse problem...Valentin De Bortoli
 
ABC convergence under well- and mis-specified models
ABC convergence under well- and mis-specified modelsABC convergence under well- and mis-specified models
ABC convergence under well- and mis-specified modelsChristian Robert
 
Delayed acceptance for Metropolis-Hastings algorithms
Delayed acceptance for Metropolis-Hastings algorithmsDelayed acceptance for Metropolis-Hastings algorithms
Delayed acceptance for Metropolis-Hastings algorithmsChristian Robert
 
NCE, GANs & VAEs (and maybe BAC)
NCE, GANs & VAEs (and maybe BAC)NCE, GANs & VAEs (and maybe BAC)
NCE, GANs & VAEs (and maybe BAC)Christian Robert
 
Unbiased Bayes for Big Data
Unbiased Bayes for Big DataUnbiased Bayes for Big Data
Unbiased Bayes for Big DataChristian Robert
 
Monte Carlo in Montréal 2017
Monte Carlo in Montréal 2017Monte Carlo in Montréal 2017
Monte Carlo in Montréal 2017Christian Robert
 
CISEA 2019: ABC consistency and convergence
CISEA 2019: ABC consistency and convergenceCISEA 2019: ABC consistency and convergence
CISEA 2019: ABC consistency and convergenceChristian Robert
 
Can we estimate a constant?
Can we estimate a constant?Can we estimate a constant?
Can we estimate a constant?Christian Robert
 
RSS discussion of Girolami and Calderhead, October 13, 2010
RSS discussion of Girolami and Calderhead, October 13, 2010RSS discussion of Girolami and Calderhead, October 13, 2010
RSS discussion of Girolami and Calderhead, October 13, 2010Christian Robert
 
ABC with Wasserstein distances
ABC with Wasserstein distancesABC with Wasserstein distances
ABC with Wasserstein distancesChristian Robert
 
accurate ABC Oliver Ratmann
accurate ABC Oliver Ratmannaccurate ABC Oliver Ratmann
accurate ABC Oliver Ratmannolli0601
 
Convergence of ABC methods
Convergence of ABC methodsConvergence of ABC methods
Convergence of ABC methodsChristian Robert
 

What's hot (20)

Mark Girolami's Read Paper 2010
Mark Girolami's Read Paper 2010Mark Girolami's Read Paper 2010
Mark Girolami's Read Paper 2010
 
Rao-Blackwellisation schemes for accelerating Metropolis-Hastings algorithms
Rao-Blackwellisation schemes for accelerating Metropolis-Hastings algorithmsRao-Blackwellisation schemes for accelerating Metropolis-Hastings algorithms
Rao-Blackwellisation schemes for accelerating Metropolis-Hastings algorithms
 
Approximate Bayesian Computation with Quasi-Likelihoods
Approximate Bayesian Computation with Quasi-LikelihoodsApproximate Bayesian Computation with Quasi-Likelihoods
Approximate Bayesian Computation with Quasi-Likelihoods
 
Bayesian hybrid variable selection under generalized linear models
Bayesian hybrid variable selection under generalized linear modelsBayesian hybrid variable selection under generalized linear models
Bayesian hybrid variable selection under generalized linear models
 
ABC-Gibbs
ABC-GibbsABC-Gibbs
ABC-Gibbs
 
Inference in generative models using the Wasserstein distance [[INI]
Inference in generative models using the Wasserstein distance [[INI]Inference in generative models using the Wasserstein distance [[INI]
Inference in generative models using the Wasserstein distance [[INI]
 
Maximum likelihood estimation of regularisation parameters in inverse problem...
Maximum likelihood estimation of regularisation parameters in inverse problem...Maximum likelihood estimation of regularisation parameters in inverse problem...
Maximum likelihood estimation of regularisation parameters in inverse problem...
 
ABC convergence under well- and mis-specified models
ABC convergence under well- and mis-specified modelsABC convergence under well- and mis-specified models
ABC convergence under well- and mis-specified models
 
Delayed acceptance for Metropolis-Hastings algorithms
Delayed acceptance for Metropolis-Hastings algorithmsDelayed acceptance for Metropolis-Hastings algorithms
Delayed acceptance for Metropolis-Hastings algorithms
 
NCE, GANs & VAEs (and maybe BAC)
NCE, GANs & VAEs (and maybe BAC)NCE, GANs & VAEs (and maybe BAC)
NCE, GANs & VAEs (and maybe BAC)
 
Unbiased Bayes for Big Data
Unbiased Bayes for Big DataUnbiased Bayes for Big Data
Unbiased Bayes for Big Data
 
Monte Carlo in Montréal 2017
Monte Carlo in Montréal 2017Monte Carlo in Montréal 2017
Monte Carlo in Montréal 2017
 
CISEA 2019: ABC consistency and convergence
CISEA 2019: ABC consistency and convergenceCISEA 2019: ABC consistency and convergence
CISEA 2019: ABC consistency and convergence
 
Can we estimate a constant?
Can we estimate a constant?Can we estimate a constant?
Can we estimate a constant?
 
RSS discussion of Girolami and Calderhead, October 13, 2010
RSS discussion of Girolami and Calderhead, October 13, 2010RSS discussion of Girolami and Calderhead, October 13, 2010
RSS discussion of Girolami and Calderhead, October 13, 2010
 
ABC with Wasserstein distances
ABC with Wasserstein distancesABC with Wasserstein distances
ABC with Wasserstein distances
 
Nested sampling
Nested samplingNested sampling
Nested sampling
 
accurate ABC Oliver Ratmann
accurate ABC Oliver Ratmannaccurate ABC Oliver Ratmann
accurate ABC Oliver Ratmann
 
Convergence of ABC methods
Convergence of ABC methodsConvergence of ABC methods
Convergence of ABC methods
 
asymptotics of ABC
asymptotics of ABCasymptotics of ABC
asymptotics of ABC
 

Similar to Poster for Bayesian Statistics in the Big Data Era conference

fb69b412-97cb-4e8d-8a28-574c09557d35-160618025920
fb69b412-97cb-4e8d-8a28-574c09557d35-160618025920fb69b412-97cb-4e8d-8a28-574c09557d35-160618025920
fb69b412-97cb-4e8d-8a28-574c09557d35-160618025920Karl Rudeen
 
The tau-leap method for simulating stochastic kinetic models
The tau-leap method for simulating stochastic kinetic modelsThe tau-leap method for simulating stochastic kinetic models
The tau-leap method for simulating stochastic kinetic modelsColin Gillespie
 
Bayesian Experimental Design for Stochastic Kinetic Models
Bayesian Experimental Design for Stochastic Kinetic ModelsBayesian Experimental Design for Stochastic Kinetic Models
Bayesian Experimental Design for Stochastic Kinetic ModelsColin Gillespie
 
DSP_FOEHU - MATLAB 02 - The Discrete-time Fourier Analysis
DSP_FOEHU - MATLAB 02 - The Discrete-time Fourier AnalysisDSP_FOEHU - MATLAB 02 - The Discrete-time Fourier Analysis
DSP_FOEHU - MATLAB 02 - The Discrete-time Fourier AnalysisAmr E. Mohamed
 
Uncertain_Henry_problem-poster.pdf
Uncertain_Henry_problem-poster.pdfUncertain_Henry_problem-poster.pdf
Uncertain_Henry_problem-poster.pdfAlexander Litvinenko
 
The moving bottleneck problem: a Hamilton-Jacobi approach
The moving bottleneck problem: a Hamilton-Jacobi approachThe moving bottleneck problem: a Hamilton-Jacobi approach
The moving bottleneck problem: a Hamilton-Jacobi approachGuillaume Costeseque
 
14th Athens Colloquium on Algorithms and Complexity (ACAC19)
14th Athens Colloquium on Algorithms and Complexity (ACAC19)14th Athens Colloquium on Algorithms and Complexity (ACAC19)
14th Athens Colloquium on Algorithms and Complexity (ACAC19)Apostolos Chalkis
 
litvinenko_Intrusion_Bari_2023.pdf
litvinenko_Intrusion_Bari_2023.pdflitvinenko_Intrusion_Bari_2023.pdf
litvinenko_Intrusion_Bari_2023.pdfAlexander Litvinenko
 
Bayesian phylogenetic inference_big4_ws_2016-10-10
Bayesian phylogenetic inference_big4_ws_2016-10-10Bayesian phylogenetic inference_big4_ws_2016-10-10
Bayesian phylogenetic inference_big4_ws_2016-10-10FredrikRonquist
 
DSP_FOEHU - MATLAB 01 - Discrete Time Signals and Systems
DSP_FOEHU - MATLAB 01 - Discrete Time Signals and SystemsDSP_FOEHU - MATLAB 01 - Discrete Time Signals and Systems
DSP_FOEHU - MATLAB 01 - Discrete Time Signals and SystemsAmr E. Mohamed
 
Research internship on optimal stochastic theory with financial application u...
Research internship on optimal stochastic theory with financial application u...Research internship on optimal stochastic theory with financial application u...
Research internship on optimal stochastic theory with financial application u...Asma Ben Slimene
 
Presentation on stochastic control problem with financial applications (Merto...
Presentation on stochastic control problem with financial applications (Merto...Presentation on stochastic control problem with financial applications (Merto...
Presentation on stochastic control problem with financial applications (Merto...Asma Ben Slimene
 
Numerical method for pricing american options under regime
Numerical method for pricing american options under regime Numerical method for pricing american options under regime
Numerical method for pricing american options under regime Alexander Decker
 

Similar to Poster for Bayesian Statistics in the Big Data Era conference (20)

fb69b412-97cb-4e8d-8a28-574c09557d35-160618025920
fb69b412-97cb-4e8d-8a28-574c09557d35-160618025920fb69b412-97cb-4e8d-8a28-574c09557d35-160618025920
fb69b412-97cb-4e8d-8a28-574c09557d35-160618025920
 
Project Paper
Project PaperProject Paper
Project Paper
 
The tau-leap method for simulating stochastic kinetic models
The tau-leap method for simulating stochastic kinetic modelsThe tau-leap method for simulating stochastic kinetic models
The tau-leap method for simulating stochastic kinetic models
 
Bayesian Experimental Design for Stochastic Kinetic Models
Bayesian Experimental Design for Stochastic Kinetic ModelsBayesian Experimental Design for Stochastic Kinetic Models
Bayesian Experimental Design for Stochastic Kinetic Models
 
DSP_FOEHU - MATLAB 02 - The Discrete-time Fourier Analysis
DSP_FOEHU - MATLAB 02 - The Discrete-time Fourier AnalysisDSP_FOEHU - MATLAB 02 - The Discrete-time Fourier Analysis
DSP_FOEHU - MATLAB 02 - The Discrete-time Fourier Analysis
 
intro
introintro
intro
 
Laplace transform
Laplace transformLaplace transform
Laplace transform
 
Uncertain_Henry_problem-poster.pdf
Uncertain_Henry_problem-poster.pdfUncertain_Henry_problem-poster.pdf
Uncertain_Henry_problem-poster.pdf
 
The moving bottleneck problem: a Hamilton-Jacobi approach
The moving bottleneck problem: a Hamilton-Jacobi approachThe moving bottleneck problem: a Hamilton-Jacobi approach
The moving bottleneck problem: a Hamilton-Jacobi approach
 
pRO
pROpRO
pRO
 
14th Athens Colloquium on Algorithms and Complexity (ACAC19)
14th Athens Colloquium on Algorithms and Complexity (ACAC19)14th Athens Colloquium on Algorithms and Complexity (ACAC19)
14th Athens Colloquium on Algorithms and Complexity (ACAC19)
 
litvinenko_Intrusion_Bari_2023.pdf
litvinenko_Intrusion_Bari_2023.pdflitvinenko_Intrusion_Bari_2023.pdf
litvinenko_Intrusion_Bari_2023.pdf
 
Bayesian phylogenetic inference_big4_ws_2016-10-10
Bayesian phylogenetic inference_big4_ws_2016-10-10Bayesian phylogenetic inference_big4_ws_2016-10-10
Bayesian phylogenetic inference_big4_ws_2016-10-10
 
HMC and NUTS
HMC and NUTSHMC and NUTS
HMC and NUTS
 
poster2
poster2poster2
poster2
 
litvinenko_Gamm2023.pdf
litvinenko_Gamm2023.pdflitvinenko_Gamm2023.pdf
litvinenko_Gamm2023.pdf
 
DSP_FOEHU - MATLAB 01 - Discrete Time Signals and Systems
DSP_FOEHU - MATLAB 01 - Discrete Time Signals and SystemsDSP_FOEHU - MATLAB 01 - Discrete Time Signals and Systems
DSP_FOEHU - MATLAB 01 - Discrete Time Signals and Systems
 
Research internship on optimal stochastic theory with financial application u...
Research internship on optimal stochastic theory with financial application u...Research internship on optimal stochastic theory with financial application u...
Research internship on optimal stochastic theory with financial application u...
 
Presentation on stochastic control problem with financial applications (Merto...
Presentation on stochastic control problem with financial applications (Merto...Presentation on stochastic control problem with financial applications (Merto...
Presentation on stochastic control problem with financial applications (Merto...
 
Numerical method for pricing american options under regime
Numerical method for pricing american options under regime Numerical method for pricing american options under regime
Numerical method for pricing american options under regime
 

More from Christian Robert

Asymptotics of ABC, lecture, Collège de France
Asymptotics of ABC, lecture, Collège de FranceAsymptotics of ABC, lecture, Collège de France
Asymptotics of ABC, lecture, Collège de FranceChristian Robert
 
Workshop in honour of Don Poskitt and Gael Martin
Workshop in honour of Don Poskitt and Gael MartinWorkshop in honour of Don Poskitt and Gael Martin
Workshop in honour of Don Poskitt and Gael MartinChristian Robert
 
How many components in a mixture?
How many components in a mixture?How many components in a mixture?
How many components in a mixture?Christian Robert
 
Testing for mixtures at BNP 13
Testing for mixtures at BNP 13Testing for mixtures at BNP 13
Testing for mixtures at BNP 13Christian Robert
 
Inferring the number of components: dream or reality?
Inferring the number of components: dream or reality?Inferring the number of components: dream or reality?
Inferring the number of components: dream or reality?Christian Robert
 
Testing for mixtures by seeking components
Testing for mixtures by seeking componentsTesting for mixtures by seeking components
Testing for mixtures by seeking componentsChristian Robert
 
discussion on Bayesian restricted likelihood
discussion on Bayesian restricted likelihooddiscussion on Bayesian restricted likelihood
discussion on Bayesian restricted likelihoodChristian Robert
 
Laplace's Demon: seminar #1
Laplace's Demon: seminar #1Laplace's Demon: seminar #1
Laplace's Demon: seminar #1Christian Robert
 
Likelihood-free Design: a discussion
Likelihood-free Design: a discussionLikelihood-free Design: a discussion
Likelihood-free Design: a discussionChristian Robert
 
a discussion of Chib, Shin, and Simoni (2017-8) Bayesian moment models
a discussion of Chib, Shin, and Simoni (2017-8) Bayesian moment modelsa discussion of Chib, Shin, and Simoni (2017-8) Bayesian moment models
a discussion of Chib, Shin, and Simoni (2017-8) Bayesian moment modelsChristian Robert
 
prior selection for mixture estimation
prior selection for mixture estimationprior selection for mixture estimation
prior selection for mixture estimationChristian Robert
 
better together? statistical learning in models made of modules
better together? statistical learning in models made of modulesbetter together? statistical learning in models made of modules
better together? statistical learning in models made of modulesChristian Robert
 
comments on exponential ergodicity of the bouncy particle sampler
comments on exponential ergodicity of the bouncy particle samplercomments on exponential ergodicity of the bouncy particle sampler
comments on exponential ergodicity of the bouncy particle samplerChristian Robert
 

More from Christian Robert (18)

Asymptotics of ABC, lecture, Collège de France
Asymptotics of ABC, lecture, Collège de FranceAsymptotics of ABC, lecture, Collège de France
Asymptotics of ABC, lecture, Collège de France
 
Workshop in honour of Don Poskitt and Gael Martin
Workshop in honour of Don Poskitt and Gael MartinWorkshop in honour of Don Poskitt and Gael Martin
Workshop in honour of Don Poskitt and Gael Martin
 
discussion of ICML23.pdf
discussion of ICML23.pdfdiscussion of ICML23.pdf
discussion of ICML23.pdf
 
How many components in a mixture?
How many components in a mixture?How many components in a mixture?
How many components in a mixture?
 
restore.pdf
restore.pdfrestore.pdf
restore.pdf
 
Testing for mixtures at BNP 13
Testing for mixtures at BNP 13Testing for mixtures at BNP 13
Testing for mixtures at BNP 13
 
Inferring the number of components: dream or reality?
Inferring the number of components: dream or reality?Inferring the number of components: dream or reality?
Inferring the number of components: dream or reality?
 
CDT 22 slides.pdf
CDT 22 slides.pdfCDT 22 slides.pdf
CDT 22 slides.pdf
 
Testing for mixtures by seeking components
Testing for mixtures by seeking componentsTesting for mixtures by seeking components
Testing for mixtures by seeking components
 
discussion on Bayesian restricted likelihood
discussion on Bayesian restricted likelihooddiscussion on Bayesian restricted likelihood
discussion on Bayesian restricted likelihood
 
eugenics and statistics
eugenics and statisticseugenics and statistics
eugenics and statistics
 
Laplace's Demon: seminar #1
Laplace's Demon: seminar #1Laplace's Demon: seminar #1
Laplace's Demon: seminar #1
 
ABC-Gibbs
ABC-GibbsABC-Gibbs
ABC-Gibbs
 
Likelihood-free Design: a discussion
Likelihood-free Design: a discussionLikelihood-free Design: a discussion
Likelihood-free Design: a discussion
 
a discussion of Chib, Shin, and Simoni (2017-8) Bayesian moment models
a discussion of Chib, Shin, and Simoni (2017-8) Bayesian moment modelsa discussion of Chib, Shin, and Simoni (2017-8) Bayesian moment models
a discussion of Chib, Shin, and Simoni (2017-8) Bayesian moment models
 
prior selection for mixture estimation
prior selection for mixture estimationprior selection for mixture estimation
prior selection for mixture estimation
 
better together? statistical learning in models made of modules
better together? statistical learning in models made of modulesbetter together? statistical learning in models made of modules
better together? statistical learning in models made of modules
 
comments on exponential ergodicity of the bouncy particle sampler
comments on exponential ergodicity of the bouncy particle samplercomments on exponential ergodicity of the bouncy particle sampler
comments on exponential ergodicity of the bouncy particle sampler
 

Recently uploaded

Botany 4th semester file By Sumit Kumar yadav.pdf
Botany 4th semester file By Sumit Kumar yadav.pdfBotany 4th semester file By Sumit Kumar yadav.pdf
Botany 4th semester file By Sumit Kumar yadav.pdfSumit Kumar yadav
 
SAMASTIPUR CALL GIRL 7857803690 LOW PRICE ESCORT SERVICE
SAMASTIPUR CALL GIRL 7857803690  LOW PRICE  ESCORT SERVICESAMASTIPUR CALL GIRL 7857803690  LOW PRICE  ESCORT SERVICE
SAMASTIPUR CALL GIRL 7857803690 LOW PRICE ESCORT SERVICEayushi9330
 
Biological Classification BioHack (3).pdf
Biological Classification BioHack (3).pdfBiological Classification BioHack (3).pdf
Biological Classification BioHack (3).pdfmuntazimhurra
 
9654467111 Call Girls In Raj Nagar Delhi Short 1500 Night 6000
9654467111 Call Girls In Raj Nagar Delhi Short 1500 Night 60009654467111 Call Girls In Raj Nagar Delhi Short 1500 Night 6000
9654467111 Call Girls In Raj Nagar Delhi Short 1500 Night 6000Sapana Sha
 
Vip profile Call Girls In Lonavala 9748763073 For Genuine Sex Service At Just...
Vip profile Call Girls In Lonavala 9748763073 For Genuine Sex Service At Just...Vip profile Call Girls In Lonavala 9748763073 For Genuine Sex Service At Just...
Vip profile Call Girls In Lonavala 9748763073 For Genuine Sex Service At Just...Monika Rani
 
Chemistry 4th semester series (krishna).pdf
Chemistry 4th semester series (krishna).pdfChemistry 4th semester series (krishna).pdf
Chemistry 4th semester series (krishna).pdfSumit Kumar yadav
 
Animal Communication- Auditory and Visual.pptx
Animal Communication- Auditory and Visual.pptxAnimal Communication- Auditory and Visual.pptx
Animal Communication- Auditory and Visual.pptxUmerFayaz5
 
Isotopic evidence of long-lived volcanism on Io
Isotopic evidence of long-lived volcanism on IoIsotopic evidence of long-lived volcanism on Io
Isotopic evidence of long-lived volcanism on IoSérgio Sacani
 
COST ESTIMATION FOR A RESEARCH PROJECT.pptx
COST ESTIMATION FOR A RESEARCH PROJECT.pptxCOST ESTIMATION FOR A RESEARCH PROJECT.pptx
COST ESTIMATION FOR A RESEARCH PROJECT.pptxFarihaAbdulRasheed
 
Bacterial Identification and Classifications
Bacterial Identification and ClassificationsBacterial Identification and Classifications
Bacterial Identification and ClassificationsAreesha Ahmad
 
PossibleEoarcheanRecordsoftheGeomagneticFieldPreservedintheIsuaSupracrustalBe...
PossibleEoarcheanRecordsoftheGeomagneticFieldPreservedintheIsuaSupracrustalBe...PossibleEoarcheanRecordsoftheGeomagneticFieldPreservedintheIsuaSupracrustalBe...
PossibleEoarcheanRecordsoftheGeomagneticFieldPreservedintheIsuaSupracrustalBe...Sérgio Sacani
 
SCIENCE-4-QUARTER4-WEEK-4-PPT-1 (1).pptx
SCIENCE-4-QUARTER4-WEEK-4-PPT-1 (1).pptxSCIENCE-4-QUARTER4-WEEK-4-PPT-1 (1).pptx
SCIENCE-4-QUARTER4-WEEK-4-PPT-1 (1).pptxRizalinePalanog2
 
Creating and Analyzing Definitive Screening Designs
Creating and Analyzing Definitive Screening DesignsCreating and Analyzing Definitive Screening Designs
Creating and Analyzing Definitive Screening DesignsNurulAfiqah307317
 
Recombinant DNA technology (Immunological screening)
Recombinant DNA technology (Immunological screening)Recombinant DNA technology (Immunological screening)
Recombinant DNA technology (Immunological screening)PraveenaKalaiselvan1
 
Nightside clouds and disequilibrium chemistry on the hot Jupiter WASP-43b
Nightside clouds and disequilibrium chemistry on the hot Jupiter WASP-43bNightside clouds and disequilibrium chemistry on the hot Jupiter WASP-43b
Nightside clouds and disequilibrium chemistry on the hot Jupiter WASP-43bSérgio Sacani
 
Pulmonary drug delivery system M.pharm -2nd sem P'ceutics
Pulmonary drug delivery system M.pharm -2nd sem P'ceuticsPulmonary drug delivery system M.pharm -2nd sem P'ceutics
Pulmonary drug delivery system M.pharm -2nd sem P'ceuticssakshisoni2385
 
High Class Escorts in Hyderabad ₹7.5k Pick Up & Drop With Cash Payment 969456...
High Class Escorts in Hyderabad ₹7.5k Pick Up & Drop With Cash Payment 969456...High Class Escorts in Hyderabad ₹7.5k Pick Up & Drop With Cash Payment 969456...
High Class Escorts in Hyderabad ₹7.5k Pick Up & Drop With Cash Payment 969456...chandars293
 
Asymmetry in the atmosphere of the ultra-hot Jupiter WASP-76 b
Asymmetry in the atmosphere of the ultra-hot Jupiter WASP-76 bAsymmetry in the atmosphere of the ultra-hot Jupiter WASP-76 b
Asymmetry in the atmosphere of the ultra-hot Jupiter WASP-76 bSérgio Sacani
 
Nanoparticles synthesis and characterization​ ​
Nanoparticles synthesis and characterization​  ​Nanoparticles synthesis and characterization​  ​
Nanoparticles synthesis and characterization​ ​kaibalyasahoo82800
 

Recently uploaded (20)

Botany 4th semester file By Sumit Kumar yadav.pdf
Botany 4th semester file By Sumit Kumar yadav.pdfBotany 4th semester file By Sumit Kumar yadav.pdf
Botany 4th semester file By Sumit Kumar yadav.pdf
 
SAMASTIPUR CALL GIRL 7857803690 LOW PRICE ESCORT SERVICE
SAMASTIPUR CALL GIRL 7857803690  LOW PRICE  ESCORT SERVICESAMASTIPUR CALL GIRL 7857803690  LOW PRICE  ESCORT SERVICE
SAMASTIPUR CALL GIRL 7857803690 LOW PRICE ESCORT SERVICE
 
Biological Classification BioHack (3).pdf
Biological Classification BioHack (3).pdfBiological Classification BioHack (3).pdf
Biological Classification BioHack (3).pdf
 
9654467111 Call Girls In Raj Nagar Delhi Short 1500 Night 6000
9654467111 Call Girls In Raj Nagar Delhi Short 1500 Night 60009654467111 Call Girls In Raj Nagar Delhi Short 1500 Night 6000
9654467111 Call Girls In Raj Nagar Delhi Short 1500 Night 6000
 
Vip profile Call Girls In Lonavala 9748763073 For Genuine Sex Service At Just...
Vip profile Call Girls In Lonavala 9748763073 For Genuine Sex Service At Just...Vip profile Call Girls In Lonavala 9748763073 For Genuine Sex Service At Just...
Vip profile Call Girls In Lonavala 9748763073 For Genuine Sex Service At Just...
 
Chemistry 4th semester series (krishna).pdf
Chemistry 4th semester series (krishna).pdfChemistry 4th semester series (krishna).pdf
Chemistry 4th semester series (krishna).pdf
 
Animal Communication- Auditory and Visual.pptx
Animal Communication- Auditory and Visual.pptxAnimal Communication- Auditory and Visual.pptx
Animal Communication- Auditory and Visual.pptx
 
CELL -Structural and Functional unit of life.pdf
CELL -Structural and Functional unit of life.pdfCELL -Structural and Functional unit of life.pdf
CELL -Structural and Functional unit of life.pdf
 
Isotopic evidence of long-lived volcanism on Io
Isotopic evidence of long-lived volcanism on IoIsotopic evidence of long-lived volcanism on Io
Isotopic evidence of long-lived volcanism on Io
 
COST ESTIMATION FOR A RESEARCH PROJECT.pptx
COST ESTIMATION FOR A RESEARCH PROJECT.pptxCOST ESTIMATION FOR A RESEARCH PROJECT.pptx
COST ESTIMATION FOR A RESEARCH PROJECT.pptx
 
Bacterial Identification and Classifications
Bacterial Identification and ClassificationsBacterial Identification and Classifications
Bacterial Identification and Classifications
 
PossibleEoarcheanRecordsoftheGeomagneticFieldPreservedintheIsuaSupracrustalBe...
PossibleEoarcheanRecordsoftheGeomagneticFieldPreservedintheIsuaSupracrustalBe...PossibleEoarcheanRecordsoftheGeomagneticFieldPreservedintheIsuaSupracrustalBe...
PossibleEoarcheanRecordsoftheGeomagneticFieldPreservedintheIsuaSupracrustalBe...
 
SCIENCE-4-QUARTER4-WEEK-4-PPT-1 (1).pptx
SCIENCE-4-QUARTER4-WEEK-4-PPT-1 (1).pptxSCIENCE-4-QUARTER4-WEEK-4-PPT-1 (1).pptx
SCIENCE-4-QUARTER4-WEEK-4-PPT-1 (1).pptx
 
Creating and Analyzing Definitive Screening Designs
Creating and Analyzing Definitive Screening DesignsCreating and Analyzing Definitive Screening Designs
Creating and Analyzing Definitive Screening Designs
 
Recombinant DNA technology (Immunological screening)
Recombinant DNA technology (Immunological screening)Recombinant DNA technology (Immunological screening)
Recombinant DNA technology (Immunological screening)
 
Nightside clouds and disequilibrium chemistry on the hot Jupiter WASP-43b
Nightside clouds and disequilibrium chemistry on the hot Jupiter WASP-43bNightside clouds and disequilibrium chemistry on the hot Jupiter WASP-43b
Nightside clouds and disequilibrium chemistry on the hot Jupiter WASP-43b
 
Pulmonary drug delivery system M.pharm -2nd sem P'ceutics
Pulmonary drug delivery system M.pharm -2nd sem P'ceuticsPulmonary drug delivery system M.pharm -2nd sem P'ceutics
Pulmonary drug delivery system M.pharm -2nd sem P'ceutics
 
High Class Escorts in Hyderabad ₹7.5k Pick Up & Drop With Cash Payment 969456...
High Class Escorts in Hyderabad ₹7.5k Pick Up & Drop With Cash Payment 969456...High Class Escorts in Hyderabad ₹7.5k Pick Up & Drop With Cash Payment 969456...
High Class Escorts in Hyderabad ₹7.5k Pick Up & Drop With Cash Payment 969456...
 
Asymmetry in the atmosphere of the ultra-hot Jupiter WASP-76 b
Asymmetry in the atmosphere of the ultra-hot Jupiter WASP-76 bAsymmetry in the atmosphere of the ultra-hot Jupiter WASP-76 b
Asymmetry in the atmosphere of the ultra-hot Jupiter WASP-76 b
 
Nanoparticles synthesis and characterization​ ​
Nanoparticles synthesis and characterization​  ​Nanoparticles synthesis and characterization​  ​
Nanoparticles synthesis and characterization​ ​
 

Poster for Bayesian Statistics in the Big Data Era conference

  • 1. Faster Hamiltonian Monte Carlo by Learning Leapfrog Scale Changye Wu, Julien Stoehr and Christian P. Robert. Université Paris-Dauphine, Université PSL, CNRS, CEREMADE, 75016 PARIS, FRANCE Abstract Hamiltonian Monte Carlo samplers have become standard algo- rithms for MCMC implementations, as opposed to more basic versions, but they still require some amount of tuning and cali- bration. Exploiting the U-turn criterion of the NUTS algorithm [2], we propose a version of HMC that relies on the distribution of the integration time of the associated leapfrog integrator. Using in addition the primal-dual averaging method for tuning the step size of the integrator, we achieve an essentially calibra- tion free version of HMC. When compared with the original NUTS on benchmarks, this algorithm exhibits a significantly improved efficiency. Hamiltonian Monte Carlo (HMC, [3]) Consider a density π on Θ ⊂ Rd with respect to the Lebesgue measure, π(θ) ∝ exp{−U(θ)}, where U ∈ C1 (Θ). Aim: generate a Markov chain (θ1, . . . , θN) with invariant dis- tribution π to estimate, for some function h, functionals with respect to π, 1 N N n=1 h(θn) a.s. −→ N→+∞ Θ h(θ)π(dθ). Principle: sample from an augmented target distribution π(θ, v) = π(θ)N(v | 0, M) ∝ exp {−H(θ, v)} . • auxiliary variable v ∈ Rd referred to as momentum variable as opposed to θ referred to as position, • marginal chain in θ is the distribution of interest. Hamiltonian dynamics: generating proposals for (θ, v) based on    dθ dt = ∂H ∂v = M−1 v dv dt = − ∂H ∂θ = − U(θ). ⊕ leaves π invariant, allows large moves, requires the solution flow to the differential equa- tions. (θ, v) (θ , v ) Leapfrog integrator: second order symplectic integrator which yields an approximate solution flow by iterating the fol- lowing procedure from (θ0, v0) = (θ, v)    r = vn − /2 U(θn), θn+1 = θn + M−1 r, vn+1 = r − /2 U(θn + 1). • : a discretisation time-step. • L: a number of leapfrog steps solution at a time t = L (θ, v) (θ , v ) This scheme does no longer leave the measure π invariant! Correction: an accept-reject step is introduced. A transition from (θ, v) to proposal θ , −v is accepted with probability ρ θ, v, θ , v = 1 ∧ exp H(θ, v) − H θ , −v . Pros & cons: ⊕ The algorithm theoretically benefits from a fast explo- ration of the parameter space by accepting large transi- tions with high probability. High sensitivity to hand-tuned parameters, namely the step size of the discretisation scheme, the number of steps L of the integrator, and the covariance matrix M. The No-U-Turn Sampler (NUTS, [2]) Idea: version of HMC sampler that eliminates the need to specify the number L by adaptively choosing the locally largest value at each iteration of the algorithm. How? Doubling the leapfrog path, either forward or back- ward with equal probability, until the backward and the for- ward end points of the path, (θ− , v− ) and (θ+ , v+ ), satisfy (θ+ − θ− ) · M−1 v− < 0 or (θ+ − θ− ) · M−1 v+ < 0. (θ, v) (θ+ , v+ ) (θ− , v− ) Proposal: sampling along the generated trajectory • slice sampling [2] • multinomial sampling ([1], version implemented in Stan). What about ? Tuned via primal-dual averaging [4], by aim- ing at a targeted acceptance probability δ0 ∈ (0, 1). Numerical experiment: Susceptible-Infected-Recovered Model (SIR) SIR: model used to represent disease transmission, for epidemics like cholera, within a population ηk ∼ Poisson(ytk−1,1 − ytk,1) and ˆBk ∼ LN(ytk,4, 0.152 ), where    k = 1, · · · , Nt = 20, tk = 7k. The dynamic of yt ∈ R4 is dyt,1 dt = − βyt,4 yt,4 + κ0 yt,1, dyt,2 dt = βyt,4 yt,4 + κ0 yt,1 − γyt,2, dyt,3 dt = γyt,3, dyt,4 dt = ξyt,2 − φyt,4, Interpretation: • yt,1, yt,2, yt,3 : no. of susceptible, infected, and recovered people within the community. • yt,4: concentration of the virus in the water reservoir. • ηk: size of the pop. that becomes infected during [tk−1, tk]. Assumptions: • the size of the population yt,1 +yt,2 +yt,3 is constant, • β ∼ C(0, 2.5), γ ∼ C(0, 1), ξ ∼ C(0, 25), φ ∼ C(0, 1), Observed dataset: https://github.com/stan-dev/stat_comp_benchmarks/tree/master/benchmarks/sir. q q qq qq qqq qqqq q qqqqq q q q qq q q q q qqqqqq q q q qq q qq q q q q qqq q qqq q qqq q qq q qq q qqqqq qq q q q q q q qq q q q q q q qq q q q q q q q q q qqq q qqq qqq q q q q qqqqqqq qqq q qqqq qq q qq q q qqq qq q qqqqqq q qqq q q q qq q qqq q qq qq qqq q q q qqqqqqq q q q qqqq q qqq q qqq q qqq qqqqqq qqqqqqqqqqqqqqqqqq q q q q qqqq q qqqq qq q qqq q q q q q q qq q q q qq q q qqq q qqqqq q q qq q q q q q qqqqq qq q q q q q qq q q q q qq q q q q q qq q qq qq q qq qq q q qqq qq qq q q q q qqq qqqq q q q q qq qqqq q qq q qq q q q q qqq qq q q q qqq q q q q q q q q qqq q qqq q q q qq q q qqq q q qqq q q q q q q q q q q q q q q q q q qq q q q q q q q q q qq q q q q q qq q q q qqq qqq qqq q qq q q qq q q q q qq q q q q q qq q q qq qq q q q q q q q q q q q qqq q qq qq q qq q q q q qqqq q q q q q q q q qq qq q q q q q q q q q q q qq q q q q q q qq q qq q q q qqq q qq q q q q q qqq q q qqq q q q q q q q q q q q q q q q q q q q q q qq q q q qq qq q q q q q q q q q q q q q qq q q q q q q qq q q q q qq q q q q q q q qq qq q q qq q q q q qqq qq q q q q q q NUTS eHMC 0.6 0.7 0.8 0.9 0.6 0.7 0.8 0.9 0.00 0.01 0.02 ESJDpergradient qq q q q q q q q qq q q qq q q qq q qqq q q q q q q q q q q q q q q q q q q q q q q q q q q qq q q q q q q qq q q qqq q q q q q q q q q q q q qq q q q q q q q q qq qq q q qqq q qq qqq qqqq q q q q q q qqqqq q q q q q q q q q q q q q q q q q q qqqqq q q q qq q q q q q q qq qq qqq q q q qq q qq qq q qq q q q q qq q q qq q qq q qqq q q qq q q q q qq q qq q qqqqq q q qq q q qqq q q qq q qq q q qq q q q qq q q q qqq q q q q q q qqq q qq q q qq q q q qqq q q qq q q q q q q qqq q q q q q q q q qq qq q q q q q q q qqq q qqq qqq q q q qq q qq q q q qq q q q q q q q qq q q q q qq q q q q q q qqqqq q q q q qq q qq q q q q q qq q q q q q q q q q qq qq q qq q q q q qqq q q q q q q q q q q q qq q q q q q q qq q q q q qqqq q q q q q q q q q q q qqqq q q q q q q qq q q q qq q q q q q q q q qq qq qq q qqqqq q q q qq q q q q q q q qq qq q q q qq q q q q q qq q qq q qq q qqq q q qqq q qqq qq q q qq q q qq qq qq q q q q q q q q q qq q qq qq q qqq q q q q q q qqq q q q q q q q q q q q q q q q q q qqq q qq q q qqq q q q q qqq q qq q q q q q q q q qq qq q q q q qqqqq q q q qq q qqq q q q q q q q q q q q qq q q q q q q q qq q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q qq q q q q q q q qq q q q q q q q q q qq q q qq q q q q qq q q q q q q q q q q q q q q q q q q q q q q q q q q q qq q q q q q qqqq q qq q q q q q q qqq q q q q q q q q q q qq q q q q q q q q q q q q q q q qq q q q q q q q q q q qq q q q q q q q q q qqq q q q qq q q q q q q q q q q q q q q q q q q q q q q q q q q qq q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q qqq q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q qq q q q q q q q q q q q q q q q q q q q q qq q q q q q q q q q q q q q q q q q q q q q q q q qq q q q q q qq q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q qq q q q q q q q q q qqq q q q q q q q q q q q q q q q q q q q qq q q q qq q q q q q q qq q q q q q q q q q qq q q q q q q q q q q q q q q q q q q q q q q q q q q qq q q q q q q q q qq qq q q q qq q q q q q q q q q q q q q q q q q q q q q q q q q q q q qq q q qq q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q qq q q qq qq q q q q q q q q q q q q q q q qqq q qq q q q qq q q qq q q q q q q q q q q q q q q q q q q q q q q q q q q qq q q q q qqqq q q q q q qqq q qq q q q q q q qq q q q q q q q q qq qq q qqq qq qq qqqqqqqq q q q q q qq q qq q q q q q q q q q q q q qqqq q q q qqqq q q q q q q q q q q q qq q q qqq q q q qqq qq qq q q q q q q qqq q q q q qq q q qqq q q qq qqq q q qq qq q qqqqq q q q q q qqq q q q q q q qq qq q q q q q qq q qqqqq q q q q q q q qq q qq q q qq q q q qqq q q qq q q qq q q qqq q q q q q q q q qqqq q q q qq q q q qq q q qqqqq q q q qq q q qqq q qq q q q q q q q q q q qq q q q qq q q qq q q q q q q q q q q qq q q qqq q q qq q q q q q q q q q q q q q q q q qqq q qq qq q q q q q q qq q q q q q qq q q q q q qqq q q q q q q q q q q q q qq q q qq q q q q qq qq q q qq q q q q qq q q q q q q q q q q q q q q q qq qqqq q q q q q q q q q q q q q q qqq q q q qq q qq q qq q q q q qqqq q q q q q q q q q qqq qq qq q q q q q q q q qq q qq qq q q q q qq q q q q q q q q q q qqq q q q q q qq qq q q q qq qq q qq q q q q q q q q q q q q q qq qqq qqq q q q qqqqq qq qq q q q q q qq q q q q q q q q q q q q q q q q q q q q q q qq q qq q q q q q q qq q q q q qqq q q q q q q q q q q q q q q qq q q q q qq q q qq q q q q q qq q q q q q q q qq qq q q q q q q q q q qq q q q q q q q q q q q q q q q q q q q q q q q q q qqq q q q q q q q qq q q q q q qq q q q q q q q q q q q q q q q q qq q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q qq q q q q q q q qq q q q q q qqq q q q q q q q q q q q qq q q q q q q q q q q q q q q qq qq q q q q q q q q qq q q q q q qq qq q q q q q q q q q q q q q q q q q q q q q q q q q q qq q q q qq q q q q q q q q q q q q q q q q q q q q q qq q q q q qqq q q q q q qq q q q q q q qq q q q q q qq q q q q q q q q q qq q q q qqq q q q q q q q q q q q q q q q q q q q q q qq q q q q q q q q q q q q q q q q qq q q q q q q qq q q q q q q q q q q q q q q q q q q q q q q q q q qq q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q qq q q q q q q q q q q q q q q q q q qq qq q q q qq q q q q q q q q q q q q q q q q q q q q q qq q q q q q q q q q q qq q q q q q q qq q q q q q q qq q q q qq q q q q q qq q q q qq q q q qq q qq q q q q q q q q q q q q q qq q q q q q q q q q q q q q q q q q q q q q q q q qq q q q q q q q q q q q q q qq qq q q q q q q q qq q q q q q q q qq qq q qq q q q q q q q q q NUTS (xi) eHMC (xi) NUTS (phi) eHMC (phi) NUTS (beta) eHMC (beta) NUTS (gamma) eHMC (gamma) 0.6 0.7 0.8 0.9 0.6 0.7 0.8 0.9 0.6 0.7 0.8 0.9 0.6 0.7 0.8 0.9 0.005 0.010 0.015 0.020 0.025 0.005 0.010 0.015 0.020 0.025 Targeted acceptance probability δ ESSpergradient Empirical HMC (eHMC, [5]) Longest batch associated with (θ, v, ): L (θ, v) = inf ∈ N (θ − θ) · M−1 v < 0 , where (θ , v ) is the value of the pair after iterations of the leapfrog integrator. Learning leapfrog scale: tuning phase with the optimised step size and an initial number of leapfrog steps L0. At each iteration, one 1. iterates L0 leapfrog steps to generate the next state of the Markov chain. 2. computes the longest batch for the current state of the chain. Output: empirical distribution of the longest batches ˆPL = 1 K K−1 k=0 δ L (θk, v(k+1) ) . eHMC: randomly pick a number of leapfrog steps accord- ing to the empirical distribution ˆPL at each iteration of HMC algorithm. ⇒ valid: the resulting transition kernel can be seen as a com- position of multiple Markov kernels attached to the same stationary distribution π. References [1] M. Betancourt A conceptual introduction to Hamiltonian Monte Carlo. arXiv preprint arXiv:1701.02434, 2017. [2] M. D. Hoffmand and A. Gelman. The No-U-Turn Sampler: adaptively setting path lengths in Hamiltonian Monte Carlo. Journal of Machine Learning Research, 15(1):1593–1623, 2014. [3] R. M. Neal. MCMC using Hamiltonian dynamics. Handbook of Markov Chain Monte Carlo, Chapter 5, Chapman & Hall / CRC Press, 2011. [4] Y. Nesterov. Primal-dual subgradient methods for convex problems. Mathematical Programming, 120(1):221–259, 2009. [5] C. Wu, J. Stoehr and C. P. Robert. Faster Hamiltonian Monte Carlo by Learning Leapfrog Scale. arXiv preprint arXiv:1810.04449v1, 2018.